4 Research issues concerning semi-urban landscapes

When summarizing the (research) backgrounds of the publications analysed for this study, there are three main ‘research issues’ that always return, in most cases even in relationship to each other. The first – and most general – research topic which can focus on semi-urban landscapes, is the construction, study, or lack of a sustainability framework for this specific land cover. Very narrowly related to this topic, because it is inherent to the sustainability story, are the research topics concerned with the environmental impacts of, and inside, semi-urban areas. To study these latter two issues, there is obviously a need for methods, tools and technical knowledge to measure and detect or identify semi-urban areas, fringes, and sprawl dynamics and its properties – these research topics being the third often encountered issue.

4.1 Environmental impacts

Land transformations associated with urban expansion can significantly affect biodiversity, energy flows, biochemical cycles, climate conditions, hydrology, soil properties at local, regional and even larger scales (Baker et al., 2002*; Breuste et al., 1998; Sukkop and Hejny, 1990).

Urban sprawl is often seen as a key sustainability problem (Davis and Schaub, 2005*; Frenkel, 2004b; Galster et al., 2001; Hasse and Lathrop, 2003*; McDonnell et al., 1997*; Schrijnen, 2000). Sprawl causes fragmentation of open spaces with well-known consequences like habitat loss, landscape degradation, etc. (Alberti, 2005*; Alig et al., 2004; Antrop, 2004; EEA, 2006*; Gardner et al., 1993; Gonzalez-Abraham et al., 2007*; Gulinck and Wagendorp, 2002; McKinney, 2002*; Nuissl et al., 2008; Savard et al., 2000*; Tischendorf and Fahrig, 2000).

Next to the fragmentation, sprawl also “consumes” land: natural and agricultural land is transformed, sometimes at high speeds, into “artificial” land covers, like residential, industrial or service areas. Not only does this result in a general decrease in agricultural and natural land, due to the competition with these types of land cover it chances their socio-economic properties, with a standard example the simplest of all indicators, the price of land (EEA, 2006*; Felsenstein, 2002; Frenkel, 2004a; Kahn, 2000; Mori, 1998; Plantinga and Miller, 2001; Ryan and Hansel Walker, 2004).

Sprawl is characterized by auto-centered, low density communities that consume large amounts of open space per capita (Davis and Schaub, 2005; Fulton et al., 2001*) and energy because of longer transport distances (Calthorpe and Fulton, 2001; Ewing et al., 2000; Gonzalez-Abraham et al., 2007; Muñiz and Galindo, 2005*).

Sprawl is known to be the major cause of visual degradation in already densely populated areas in, for example, Southeastern Asia (Madhavan et al., 2001), California (Atkinson and Oleson, 1996; Fulton et al., 2001) or Northwestern Europe (Antrop and Van Eetvelde, 2000; EEA, 2006; Hagoort et al., 2002; Holden and Turner, 1997). Another negative effect of the existence of semi-urban areas, especially in more arid climates like the Californian or Australian deserts, is the increased fire risk (Syphard et al., 2007).

Other examples of ecological impacts of sprawl, semi-urban areas, or fringe dynamics are climate impacts. McDonnell et al. (1997*) also studied ecosystem temperatures along a rural–urban gradient, Jenerette et al. (2007) proved regional relationships between surface temperature and vegetation in urbanizing ecosystems, Baker et al. (2002) studied this for the Phoenix metropolitan area in Arizona and Saaroni et al. (2000) studied the heat island effect in Tel-Aviv, Israel.

Figure 4: Urban sprawl is accompanied by an increase in impervious surfaces with well-known environmental impacts.

The bulk part of artificial land cover is impervious or at least only partly permeable. An increase in urban fabric automatically means a – sometimes dramatic – increase in area of impervious surface (Foley et al., 2005; Stone Jr, 2004), having impacts on all kinds of hydrological properties, like run-off and polluent infiltration (Bronstert et al., 2002; Carlson and Arthur, 2000; Haase and Nuissl, 2007; Niehoff et al., 2002; Pauleit and Duhme, 2000; Perry and Nawaz, 2008; Pickett et al., 2001*; Ziegler et al., 2004). Measuring surface sealing is an important issue when studying flood control and urban hydrology (Jennings and Jarnagin, 2002; Jennings et al., 2004*). It is no surprise that the evolution of sprawl and semi-urban areas is narrowly followed by hydrologists (Gillies et al., 2003; Hall, 1989; Hoggan, 1989; Pauleit et al., 2005; Shaw, 1988; Zheng and Baetz, 1999).

A meanwhile very common way to calculate the impact of any form of human activity, thus including urban expansion like sprawl or the existence of semi-urban areas, is by calculating the ecological footprint (Bugliarello, 2004b; Muñiz and Galindo, 2005). Rees (1992) calculated the footprint of sprawl in a valley in Canada, Alfsen-Norodom (2004*) did similar experiments for the New York metropolitan area, Doughty and Hammond (2004) did a thorough footprint analysis for the Bath region, U.K.

Another big impact of sprawl, semi-urban areas and fringe expansion and dynamics that needs to be mentioned is what can be called the ‘garden’ effect. Urban sprawl in most cases means a large growth in residential area (compared with the increase of industrial or other artificial, anthropogenic land cover). In developed countries, an increase in residential area, mostly also means an increase in garden area, since the majority of the ‘growing’ residential area, in the fringe, sprawling, in the semi-urban area, is low density housing, accompanied by gardens and on top of that public or semi-public green areas like road verges, parks, golf courses and other forms of green recreation sites (Colding, 2007; Loram et al., 2007*; Poole, 1993; Smith et al., 2005*). While, on the one hand, this increase in green biomass could be an improvement when considering it as pure ‘biomass’ and carbon dioxide storage (Bjerke et al., 2006*; McDonnell et al., 1997*; Platt, 2004; Yokohari et al., 2000), this is not always the case when discussing biodiversity. Gardens and parks possess large amounts of exotic plant species and thus not only introduce non-native species, but also increase competition between exotic and native plant species, in some cases resulting in a decrease of native flora as well as native fauna ecologically linked to those native species (Alberti, 2005; Gallent et al., 2004; McKinney, 2002*; Niemelä, 1999*; Pauchard et al., 2006; Savard et al., 2000; Theobald, 2004; Vogtländer et al., 2004). Extensive research has been done on biodiversity in and around gardens in the Urban Domestic Gardens research in, for example, the Sheffield region, U.K. (Gaston et al., 2005a,b; Loram et al., 2007; Smith et al., 2005, 2006; Thompson et al., 2003, 2005, 2004).

The important role of public green in the urban ecology debate for sustainable cities can be found in studies of biodiversity in parks (Cornelis and Hermy, 2004; Hermy and Cornelis, 2000) and their role as buffers, social important locations, and landscape connectors (Angel et al., 2005; Bjerke et al., 2006; Countryside Agency Research Programme, 2002*; Chiesura, 2004; Niemelä, 1999).

4.2 Detecting and measuring fringe, sprawl and semi-urban landscapes

Definitions or descriptions given to specific forms of land use or its dynamics, define the way it is measured. For example, large studies are conducted to measure sprawl and its impacts. Examples of studies only measuring the physical and landscape ecological properties of sprawl are done, for example, by Malpezzi and Guo² or Torrens and Alberti (2000), and the study of sprawl as a process by Wolman et al. (2005). Studies measuring the impacts of sprawl, semi-urban landscapes, and land-use interactions on the urban fringes form a long list, of which examples are discussed further on in this paper.

A specific method of studying urban fringe systems and sprawl, which is related to urban cores, is gradient analysis. Good examples of this method are the studies of McDonnell (McDonnell and Pickett, 1990; McDonnell et al., 1997), the detailed rural–urban gradient analysis of the area around Phoenix by Luck and Wu (2002), and the GIS based gradient analysis of the urban landscape and sprawl dynamics in China, in Shanghai (Zhang et al., 2004) or Guangzhou (Yu and Ng, 2007). Other rural–urban gradient studies cover ecological indicators (McKinney, 2002), residential building patterns (Weng, 2007), land use properties like vegetation type (Zhao et al., 2007), sealed surfaces (Jennings et al., 2004), or building densities (Longley and Mesev, 2002).

Figure 5: Panorama of a semi-urban landscape seen from the hills of Sintra, north of Lisbon, Portugal.

Multiple authors emphasize the importance of choosing the right scale properties when studying sprawl phenomena, the rural-urban fringe, and semi-urban areas. Especially when measuring physical geographical properties like landscape metrics as fractal dimension, diversity indices or patch densities, properties as resolution, study area extent, and classification type have great influence on the results of the measurements (Cain et al., 1997; Hasse and Lathrop, 2003; Lam and Quattrochi, 1992; O’Neill et al., 1996; Saura, 2004; Wu, 2004).

Because of the heterogeneity of semi-urban land use and the hybrid, intricate and apparently ‘unorganized’ structure, working with high resolution data for high levels of detail is often recommended both for monitoring (Cablk and Minor, 2003; Chen and Stow, 2003; Herold et al., 2002), as for modeling (Pontius Jr et al., 2004). Spectral unmixing may provide some solution for the detail enhancement of low resolution spatial data (Raymaekers et al., 2005; Wu and Murray, 2003), however, extra and more detailed ground data is needed for verification (Jensen and Cowen, 1999).

4.3 Sustainability and sprawl, semi-urban areas and urban fringes

Theobald (2001) recommends researchers to recognize the ubiquity of exurban areas and better incorporation of fine scale patterns of land use beyond the urban fringe. The literature concerning sustainability and semi-urban areas reveals three groups. The first group relates different demands for land use to each other, not only at local level (direct competition for land and incompatibility of land uses) but also at global level (footprinting). The second group rather focuses on the local to global environmental impacts other than related to land take. The third group assumes a more positive stance in focusing on the benefits and ecological opportunities of the green component in (semi-)urban areas.

Many authors suggest planning concepts or tools, derived form these theories, for sustainable urban planning. Tools like the “urban structural unit” principle (Böhm, 1998), conurbation measures (Countryside Agency Research Programme, 2002) or planning concpepts like the Biosphere project (Alfsen-Norodom, 2004; Alfsen-Norodom et al., 2004), multifunctional agriculture (Casaux et al., 2007) or neo-rurality (Gulinck, 2004) are certainly worth mentioning and open interesting paths to a general improvement of landscapes consisting of a rural–urban matrix.

Pickett et al. (2004) proposes the concept of “cities of resilience”, with a deep link between urban design and ecology, a parallel can be found in the “nurtured landscape” view of Yang and Lay (2004).

Good examples of decision support strategies as a tool for sustainable urban planning of semi-urban areas include sprawl measures, master plans, neighborhood development tools, hierarchical processes for evaluation, growth management, and the implementation of green structures as an instrument and are abundant in the literature (Banai, 2005; Cadieux, 2008; Daniels and Lapping, 2005; Hümmeler, 1998; Jensen et al., 2000; Yli-Viikari et al., 2007).